Code Evaluation on LCB v6
87.7ScoreGPT-5.2
Evaluation Results
| Method | Links | |
|---|---|---|
| GPT-5.22026.03 | 87.7 | |
| Gemini-3-Pro2026.03 | 86.9 | |
| Kimi-K2.5Number of Parameters=1T-A32B2026.03 | 85 | |
| Intern-S1-ProNumber of Parameters=1T-A22B2026.03 | 74.3 | |
| Qwen3-VL-235B-ThinkingNumber of Parameters=235B-A22B, Thinking Configuration=true2026.03 | 72 | |
| Nemotron-Nano-v2# Param.=9B2026.02 | 60 | |
| Falcon-H1R# Param.=7B2026.02 | 57.71 | |
| Ministral-3-R# Param.=8B2026.02 | 53.71 | |
| MiniCPM-SALA# Param.=9B2026.02 | 52 | |
| MiniCPM-4.1# Param.=8B2026.02 | 51.43 | |
| Qwen3.5-4B (released)Model Size=4B, Training Protocol=Released, Decoding Strategy=Native AR decoding2026.06 | 50.86 | |
| Qwen3.5-9B (released)Model Size=9B, Training Protocol=Released, Decoding Strategy=Native AR decoding2026.06 | 49.71 | |
| FLARE-9BModel Size=9B, Training Protocol=FLARE, Decoding Strategy=AR-Trust sampling2026.06 | 49.71 | |
| Qwen3# Param.=8B2026.02 | 48.57 | |
| Qwen3.5-9B + AR-SFTModel Size=9B, Training Protocol=AR-SFT, Decoding Strategy=Native AR decoding2026.06 | 46.29 | |
| Qwen3.5-4B + AR-SFTModel Size=4B, Training Protocol=AR-SFT, Decoding Strategy=Native AR decoding2026.06 | 45.71 | |
| FLARE-4BModel Size=4B, Training Protocol=FLARE, Decoding Strategy=AR-Trust sampling2026.06 | 41.71 | |
| Qwen3.5-2B + AR-SFTModel Size=2B, Training Protocol=AR-SFT, Decoding Strategy=Native AR decoding2026.06 | 21.14 | |
| Qwen3.5-2B (released)Model Size=2B, Training Protocol=Released, Decoding Strategy=Native AR decoding2026.06 | 17.71 | |
| FLARE-2BModel Size=2B, Training Protocol=FLARE, Decoding Strategy=AR-Trust sampling2026.06 | 15.43 |